plant metabolomics: addressing the scientific
TRANSCRIPT
®Plant metabolomics: Addressing the scientific concerns and uncertainties of genetic modificationDerek Stewart1, Ian Colquhoun3, Jim McNicol2 and Howard Davies1
Quality Health and Nutrition Programme1 & BioSS2, Scottish Crop Research Institute, Dundee, DD2 5DA3Institute of Food Research, Norwich Research Park, Colney, Norwich NR4 7UA,
Conclusion
Background
Acknowledgement
Reference
Plant Metabolomics: The discrete biological experimental level
Unintended effects; Is it the GM or something more basic?
Plant Metabolomics: The discrete biological experimental level
Approaches & Results
Shepherd, T., Dobson, G., Marshall, R., Verrall, S.R., Conner, S., Griffiths, D.W., Stewart, D. & Davies, H.V. 2005. Profiling of metabolites and volatile flavour compounds from Solanum
species using gas chromatography-mass spectrometry. Proceedings of the Third International Congress on Plant Metabolomics, Ames, Iowa, USA, June 2004
The work presented here was funded by the Food Standards Agency G02 Programme; Transcriptome, proteome
and metabolome analysis to detect unintended effects in genetically modified potato.
A combined metabolomic and statistical approach clearly functions as a high throughput method for establishing quantitative and qualitative chemical differences, in this case, with regard to unintended changes accompanying genetic modification.
Broadly speaking the majority of phytochemical variation accompanying GM was encompassed within natural biological variation
The genetic modification of plants has been shown repeated to be successful at altering target (bio)chemical components. However, concerns over the specificity of these modifications, especially when undertaken in food crops, have fuelled the public GM debate. Subsequently the more general question has arisen of whether the existing, generally targeted, safety assessments applied to traditionally bred crops are sufficient or are new strategies required? Targeted analyses will, by definition, miss unexpected or unintended compositional changes and the application of ‘catch all’ analytical technologies such as LC-MSn, GC-TOF-MSn and NMR, in toto metabolomics technologies, have been proposed as the potential next wave of safety assessment approaches.
Potato, the fourth largest global food crop, has been the subject of many specific genetic modifications and served here as an excellent model to assess the relative efficiencies of metabolomic technologies in determining unintended effects, or deviations from substantial equivalence. Several transgenic lines and their tissue culture and vector-only controls were subject to metabolomics. In addition, and of equal importance, a broad range of potato varieties and landraces (lines which have not been subjected to controlled introgression of traits from a variety of wild species) were subject to metabolomics. The data from this allowed metabolite changes accompanying genetic modification, both intended and unintended, to be reported upon within a broader scale of germplasm biodiversity.
To get a better understanding of the incidence and characterisation of unintended effects a well characterised set of GMs (and their appropriate vector-only, tissue culture and wild type controls) were analysed alongside a wide range of non GM germplasm (varieties, landraces). Standard operating procedures were developed for field experimentation, harvest sampling and extraction and derivatisation1. The extracts were analysed by LC-MSn, GC-ToF-MS and NMR and that data subject to uni- and multivariate statistical analysis to identify specific differences and patterns in the data
35S-SAMDCPatatin-SAMDCGBSS prom.W2 geneMal 1 35SGBSS prom.SGTFructokinase
Extremely stunted phenotypeTuber numbers increasedWaxy starch phenotypeGlycoprotein processing – stunted growth50% reduction glycoalkaloidsStarch Sugar metabolism
Regulation
GM lines
Gene Comments
Cultivated “introgressed” tuberosum (Varieties) Non-introgressed (landraces)
DesireeP.DellRecordSheelaghStirlingTorridonGlennaMorag
EdenM.PiperP.JavelinCaraP. Crown BrodaBarbaraG Wonder
Inca SunMayan GoldLumpersForty FoldAnyaBrodickPF Apple91MT46E15(1)
S. tuberosum, Tetraploid
Wild Andean TBR3369 (5)TBR5646 (4)TBR3302 (1)
S. Phureja, diploid
Varieties and Landraces
Broadly, much of the metabolite variation falls within the range covered by natural variation across the three 3 analytical approaches. For LC-MSn there appears to be more of a division of the GM, vector only and tissue culture lines away from the range described by the varieties and landraces giving an indication of the chemistries being described here (Figs 1,2 &3)
Genetic modification and the apparently innocuous process of tissue culture both impact upon glycoalkaloid levels and the ratios of α-chaconine and α-solanine. Both processes decrease the total glycoalkaloid content whilst increasing the relative level of the reportedly more toxic α-chaconine. If this effect is repeated for other unrelated compounds it raises the question of how comparable plants are following repeated passage through tissue culture (Fig 4)
Broad brush metabolomics can generate impenetrable datasets which often yield interesting data once refined to a defined experimental level. For example the sense and antisense modification of S-adenosylmethionine decarboxylase in potato produced both predictable and unintended effects with increases in methione, and tyrosine and caffeoyl putescine respectively (Fig 5)
-4 -2 0 2 4
-2
-1
0
1
2
PC2
PC3
NMR Gc ToF-MSn
-4 -2 0 2 4
-2
-1
0
1
2
PC1
PC2
LC/MS
-4 -2 0 2 4
-2
-1
0
1
2
PC1
PC1 PC1 PC1
PC2
ANOVA analysis ofNMR data (tyrosine)
1.0
2.0
Inte
nsity
Inte
nsity
PCA analysis of LC-
PC1
PCA analysis of GC-ToF-MSn data (Phenylalanine)
PC1
Inte
nsity
MSn data (Caffeoyl putrescine)Global Metabolomic dataset
PC2
PC20.1
C
haco
nine
leve
l
Tota
l gly
coal
kalo
ids
Cha
coni
ne /
Sol
anin
e
20-2-4-6-820-2-4-6-820-2-4-6-8
40
30
20
10
4
2
3
1
60
50
40
30
20
10
Cultivars, landraces & controlsGMsVector onlyTissue culture
Cultivars, landraces & controlsGMsVector onlyTissue culture
Wild TypeVector onlyTissue cultureSense GM 1Sense GM 2Antisense GM
Figure 5
Figure 4
Figure 1 Figure 2 Figure 3